🔒 IRR’s snap poll – Gareth van Onselen on what it means for 2019

JOHANNESBURG — The Institute of Race Relation’s latest snap poll on the build-up to the 2019 election has drawn a lot of interest. It’s been lauded for being among the most credible yet and paints a picture of what could happen to the vote in 2019. While the ANC has been gaining lost ground, the DA seriously risks either not growing or going backwards. In the meantime, the EFF is still on a small base of just 10%-11%, but if it achieves this in 2019 it would have almost doubled its support base in 2019. In this fascinating interview, the IRR’s Gareth van Onselen explains what this snap poll means and why a lot could still change in months to come. – Gareth van Zyl

It’s a warm welcome to Gareth van Onselen from the Institute of Race Relations (IRR) in Johannesburg on the BizNews podcast. Gareth, thanks for joining us.

It’s a pleasure.


You’ve released a snap political poll this week, which showed that the ANC is gaining back some lost ground, while the DA and EFF, from my perspective at least, seem to be going nowhere in terms of national support. Can you tell us more about the exact results of this snap poll and what your methodology was?

This is the second poll that we’ve done in the last few months. We aim to do one full survey every quarter. We’ve called this one a snap poll because it falls between two quarterly surveys and we just thought it important to test the market before the end of the year. So, it’s not a full survey. We just focused on party political support. It came out of the field on 4 December. It included around 1,000 registered voters as respondents, and it was a demographically representative sample, i.e., weighted against age, gender, race, geographic location etc.

What it shows is that compared to our September poll, the ANC is up four points. So, it came out on 52% in September, and we’ve got it on 56% at the moment. It shows a significant decline for the DA, which is down from 23% in September to 18% now. And then there’s a slightly smaller decline for the EFF, down from 13 to 11 points. But just to be clear, even though the DA and the EFF have both declined that’s not necessarily a bad result for the EFF. You must remember, the EFF attained 6.3% in 2014 so, to be on 11% is almost having doubled its support and that’s actually quite a good position compared to the DA, which is now substantially behind the 22% it got in 2014. The DA’s got a real fight on its hands.

I thought it interesting that you also look at different turnout scenarios. One turnout scenario being 69%, which to me looks like it could be a very probable turnout considering that in 2014 the turnout stood at 73% and in 2009 it was at 77%. So, there’s a different result if the turnout is 69%, right?

Well, let me just explain how turnout works for your listeners. With any poll that you undertake, the results presume a 100% turnout because, obviously, the views of every single person you spoke to in the poll are recorded. But on election day not everybody goes to the polls. And as you said, in 2014 the turnout was 73% — this means 27% of voters didn’t pitch up to vote. To be realistic, you have to adjust those figures for different turnout scenarios to actually get an idea of what it would look like on the day. And like you, I think 69% is a realistic turnout scenario — it’s in the ballpark. I think the turnout will drop a bit from 2014 and 69% is there or thereabouts.

We, therefore, attach certain other questions to the survey to gauge how likely people are to vote, and then you take those people that indicate they are unlikely to vote or won’t vote out of the sample, and you see what it does to the numbers. So, on 69% the ANC jumped up even more, up to 59%. The DA recovers some ground and gets back up to 22%. So, it’s on par with where it was in 2014. The EFF loses 1%, down to 10%, which is not a bad result after all things considered, but the EFF does seem to be in the decline at the moment from its high point in September.

In South Africa, we call it collision government. More magic available at www.zapiro.com.

So, whichever way you view these results, the ANC looks like it’s going to come out tops again. Obviously, the ANC has dominated politics in SA for the last quarter of a century, and I guess that Cyril Ramaphosa coming into power would maybe attract some voters who would have shunned it previously in years gone by, such as in 2016. But it doesn’t look like the ANC is losing power any time soon, does it?

No, it doesn’t. And I think that there are two factors at play here. The one is this idea of some kind of renewal inside the ANC under Cyril Ramaphosa. And whether you agree with that or not – that sentiment seems to be very powerful out there in the public and it is no doubt responsible for the party being able to win back some of those voters, as you said, who were alienated from it under Jacob Zuma. But the other thing to bear in mind is that the ANC does disproportionately well in national elections historically. In local government elections, it always loses a huge percentage of its voters who don’t turn out. They couldn’t care less about local government elections. In fact, there were about more than two million voters who didn’t pitch up for the ANC in 2016, and that didn’t mean they switched parties. They just didn’t come to vote. But they do come to vote in national elections and that drives up its percentage. I think those two things come together to ensure that the ANC, as things stand, looks set on getting a majority.

Mmusi Maimane, DA leader
Mmusi Maimane is leader of the opposition Democratic Alliance party. Photographer: Waldo Swiegers/Bloomberg

What about the DA because if your poll manifests itself in any way next year in the actual result, it would be the first time that I can remember that the DA would not have grown, or would have even gone backwards. That, to me, looks like a possible crisis within the party there?

Yes, it is and there are a number of scenarios for the DA, ahead of the elections, and there are a number of factors you have to bear in mind as well. The one is we’ve got a whole election campaign to come and the DA has got quite a lot of money and a big election machine. If it can turn the tide a bit and grow by a few percentages on a lower turnout, then there is a real possibility that it could grow by a percent or two. But just as real, as you intonate, there’s a possibility of it not growing or even shrinking somewhat. That would be an entirely new experience for the DA. It has never failed to grow in a national election before and the kind of internal crisis that would trigger will be without precedent. It will be quite interesting to see what happens. I suspect a number of heads will be on the chopping block, whether they’re high profile figures, like leaders, or the strategic management of the party behind the scenes, I’m not sure — but people will want someone to account for the fact that they have not grown.

Let’s talk about the EFF as well because, as you’ve mentioned, they would have almost doubled their voter base, which is quite incredible. But at 10% or 11%, they’re still not clearly the choice of the majority of South African voters. So, they still appear to be very much on the fringes. Do you think that that’s a correct assessment there?

Julius Malema EFF
Julius Malema, the firebrand leader of South Africa’s EFF. REUTERS/Siphiwe Sibeko.

Yes and no. I mean, I agree with the fact that electorally 10% is by no means a dominant share of the market. But they do punch above their weight, and given the extent to which they’ve managed to define South African voters with just 6%, who knows what they’re capable of with 10% or so. They’d also have a bigger influence in the provinces, naturally, as a consequence of that, and so they’d have more ability to define the national debate. But the thing about a poll is that obviously, it’s a moment in time. So, this is how things did on 4 December, but you don’t necessarily know which direction the graph is headed in the long term. We know that there’s a short-term decline for the DA and for the EFF, whether the DA is going to peter out or turn in the other direction, the same applies to the EFF, we don’t know.

If they continue to decline and things like the VBS scandal have played into that, they could drop below 10%, and there’s every possibility [of that]… There’re a number of scenarios for the EFF, just like with the DA, where they could end up with only 1% or 2% growth if things go horribly wrong. So, all these options are on the table, I know that’s not very helpful for listeners who want absolute certainty but there are five months to go to an election and anything can change.

Lastly, this snap poll, as you’ve said, is just one of many that your organisation is doing but why is it important for the Institute of Race Relations to do polls like this? Because I’ve always associated you guys with being a politically agnostic think tank in a way — so why then do political polls?

Well, the institute offers a number of services to the public. Obviously, our primary mandate is to advocate liberal thoughts and ideas and stand up for the Constitution and things that are attached to that. But we also have a great deal of credibility and capital invested in being able to analyse current affairs and politics. I think there is a dearth of high-quality political research, which we want to fill and become authoritative on. I think you’ve got to be careful with polls as you do with any piece of analysis or opinion-making in the public mind. You have to be critical in how you review them and understand what the limitations are. But within that, I think they are a very helpful tool for understanding, both the sentiment and we don’t have enough of that kind of thing.

Gareth van Onselen, thank you so much for taking the time to chat with me today.

Not at all, it’s a pleasure.

Visited 49 times, 1 visit(s) today